date: Fri Dec 10 16:37:37 2004
from: Phil Jones
subject: Re: Land-sea differences
to: "Tett, Simon" , Harry Harris , "Kennedy, John" , "Brohan, Philip" , "Folland, Chris"
Simon,
There is a lot of spatial consistency in your map with coloured crosses and
a few diamonds. Just selecting those stations near sea level is easy to do, but is
there a simple way of selecting those sites that are nearest the sea. This
is likely much harder. All the UK stations fell off the <5m list (even Plymouth,
which we know is almost next to the sea). Just wondering if there is another
way we can get coastal stations. 0.1K is similar to 0.125 though which
suggests that however you subsample the full set you will get a number about
0.1K.
David Parker has produce a Table for AR4 which gives trends over 1979-2003 and
other periods. Here NH land warms at 0.353K/dec and SST at 0.177K/dec. For the SH,
the figures are 0.164 and 0.093 K/dec. The differences can be estimated, but not quite
for your period. I guess, I'm saying that unless you get really coastal stations can you
really say anything? Land is warming faster, so coastal land should probably be
warming slightly faster than coastal SST. Your map is differences, but does this
manifest itself as a trend?
One final thing, can you calculate for the same SST boxes, the differences between
SST and NMAT. Is this zero, or does it approach 0.1K over the 21 years?
The blue crosses over the Great Lakes and Canadian Maritimes struck me. I'm presuming
there are no issues with 10% missing data with sea ice. If there were this would make the
difference larger than 0.1?
I would like to believe that 2) is the likely answer. I ruled out 3) !
Cheers
Phil
At 10:29 10/12/2004, Tett, Simon wrote:
All,
I've managed to compute differences between station temperatures
and SST. I have done this by computing anomalous temperatures for
1981-2000 from Phil/Harry's new land records. For SST I used MOHSST (but
I've also done the calculation for the new ICOADS). There appears to be
a systematic difference between the two. Attached is a PDF file. Page 1
shows anomalous SST and station temperatures. Page 2 shows difference
between stn mean temp and SST (both ~ 20 year averages). Diamonds are
where land/SST agree to +/- 0.05K. I have been strict -- I only make the
comparison where I have < 10% missing data (monthly averages).
The simple avg over all stn - SST is about 0.1K.
Looking at the 140 stations with a height below 5m ("low lying" -- see
attached) suggests there is a systematic difference for those stations.
The mean difference for "low lying" stations is 0.125 with a 2 SE of
0.055.
The plots don't suggest that there are systematic differences that vary
as you get further away from the ocean. (Thus the look at low lying
stations). So I think we may be seeing something else! There seems to be
three possibilities:
1) We have changing bias in the land record for the period 1980-2000
leading to an over-estimate of the warming
2) The ocean warming has been underestimated (due to a change in the
observing system) -- John K. has some evidence that this is the case.
3) My code is in error....
If 2) is correct note we will be increasing global warming over the last
20 years by about 0.05K/decade or so. If 1) is correct then we will be
reducing global-warming by 0.025K/decade (less land than ocean). If 3)
then I just look silly.
Simon
Dr Simon Tett Managing Scientist, Data development and applications.
Met Office Hadley Centre (Reading Unit)
Meteorology Building, University of Reading Reading RG6 6BB
Tel: +44 (0)118 378 5614 Fax +44 (0)118 378 5615
Mobex: +44-(0)1392 886886
E-mail: simon.tett@metoffice.gov.uk [1]http://www.metoffice.gov.uk
Global climate data sets are available from [2]http://www.hadobs.org
Prof. Phil Jones
Climatic Research Unit Telephone +44 (0) 1603 592090
School of Environmental Sciences Fax +44 (0) 1603 507784
University of East Anglia
Norwich Email p.jones@uea.ac.uk
NR4 7TJ
UK
----------------------------------------------------------------------------